Measurement of bone density; bias and variation due to radiographic and photometric techniques.

"Bone density is an important and little-investigated variable about which we should like to know more" (American Journal of Physical Anthropology, 1954). Here an anthropologist echoed a nutritionist (Nutrition Reviews, 1952) regarding photoelectric densitometer estimates of x-ray opacity expressed as thickness of an aluminium or ivory step-wedge (or bone slabs) exposed on the same film as a living subject's hand or other region. This method has been used to measure bone changes in arthritis (Bywaters, 1948; Engstrom, Wegstedt, and Welin, 1948; Engstrom and Welin, 1949) and other diseases (Mortimer, Levene, and Rowe, 1937). Its most extensive development (Brown and Birtley, 1951; Mack, 1950; Mack, Brown, and Trapp, 1949; Mack, O'Brien, Smith, and Bauman, 1939; Mack, Trapp, and Brown, 1950) has culminated in the formation of a Bone Density Centre (Brown, 1953; McFarland, 1954) to which films can be sent for rapid assessment by an elaborate and ingenious equipment, and the description of this technique has been so convincing as to prompt the suggestion that it might take the place of calcium balance analyses (Hoobler, 1950). Many radiologists and rheumatologists, however, remain unconvinced, chiefly because of the variation produced in films by x-raying and processing, and the masking of bone mineral loss by overlying soft tissue (and by internal organic matter). These problems arose in two current studies in this department: a search for age differences in bone density in healthy adults, and a seriatim comparison of densities in rheumatoid arthritics, some treated by